Load Data from the Clock Calculator

## We have 132 observations with 102 columns.

Duplicated Observations

We have 3 subjects (SIDs: CQ 1328, CQ 1424, CQ 1518) with duplicated observations.

Clocks stable?

In section, we check if the clocks are stable by comparing them in the duplicated observations.

Concordance between pairs

Evaluate correlation/concordance between the duplicate pairs

Absolute differences in two observations

Clocks / Age Accelerations

##      Horvath EAA Hannum EAA PhenoAge EAA Skin&Blood EAA GrimAge EAA     DNAmTL
## [1,]  0.05638483   2.678389     3.898440      0.6858806   0.7138976 0.04309056
## [2,]  2.80993321   1.122348     2.464987      0.8001586   1.6632470 0.04052281
## [3,]  1.23251949   0.732867     7.594791      2.9191094   1.0295776 0.02069328
##           IEAA     EEAA
## [1,] 0.3889968 3.667739
## [2,] 3.3026780 1.740232
## [3,] 0.9771748 3.076599

## 
## Summary: :

##      Horvath EAA Hannum EAA PhenoAge EAA Skin&Blood EAA GrimAge EAA     DNAmTL
## mean    1.366279   1.511201     4.652739       1.468383   1.1355741 0.03476888
## sd      1.381639   1.029403     2.646780       1.257665   0.4834692 0.01225725
##          IEAA      EEAA
## mean 1.556283 2.8281897
## sd   1.540750 0.9874722

ICC of each clock

## 
##  Horvath EAA : 
##                          type       ICC        F df1 df2           p
## Single_raters_absolute   ICC1 0.9802908 100.4755   2   3 0.001783993
## Single_random_raters     ICC2 0.9803666 165.2373   2   2 0.006015498
## Single_fixed_raters      ICC3 0.9879690 165.2373   2   2 0.006015498
## Average_raters_absolute ICC1k 0.9900473 100.4755   2   3 0.001783993
## Average_random_raters   ICC2k 0.9900860 165.2373   2   2 0.006015498
## Average_fixed_raters    ICC3k 0.9939481 165.2373   2   2 0.006015498
## 
##  Hannum EAA : 
##                          type       ICC        F df1 df2          p
## Single_raters_absolute   ICC1 0.9241935 25.38296   2   3 0.01318018
## Single_random_raters     ICC2 0.9242641 26.02300   2   2 0.03700552
## Single_fixed_raters      ICC3 0.9259890 26.02300   2   2 0.03700552
## Average_raters_absolute ICC1k 0.9606035 25.38296   2   3 0.01318018
## Average_random_raters   ICC2k 0.9606416 26.02300   2   2 0.03700552
## Average_fixed_raters    ICC3k 0.9615725 26.02300   2   2 0.03700552
## 
##  PhenoAge EAA : 
##                          type       ICC        F df1 df2         p
## Single_raters_absolute   ICC1 0.3206087 1.943812   2   3 0.2874602
## Single_random_raters     ICC2 0.3206087 1.943812   2   2 0.3396956
## Single_fixed_raters      ICC3 0.3206087 1.943812   2   2 0.3396956
## Average_raters_absolute ICC1k 0.4855469 1.943812   2   3 0.2874602
## Average_random_raters   ICC2k 0.4855469 1.943812   2   2 0.3396956
## Average_fixed_raters    ICC3k 0.4855469 1.943812   2   2 0.3396956
## 
##  Skin&Blood EAA : 
##                          type       ICC       F df1 df2         p
## Single_raters_absolute   ICC1 0.3135357 1.91348   2   3 0.2913003
## Single_random_raters     ICC2 0.3135357 1.91348   2   2 0.3432322
## Single_fixed_raters      ICC3 0.3135357 1.91348   2   2 0.3432322
## Average_raters_absolute ICC1k 0.4773919 1.91348   2   3 0.2913003
## Average_random_raters   ICC2k 0.4773919 1.91348   2   2 0.3432322
## Average_fixed_raters    ICC3k 0.4773919 1.91348   2   2 0.3432322
## 
##  GrimAge EAA : 
##                          type       ICC        F df1 df2           p
## Single_raters_absolute   ICC1 0.9807695 103.0017   2   3 0.001719697
## Single_random_raters     ICC2 0.9809233 636.9022   2   2 0.001567638
## Single_fixed_raters      ICC3 0.9968647 636.9022   2   2 0.001567638
## Average_raters_absolute ICC1k 0.9902914 103.0017   2   3 0.001719697
## Average_random_raters   ICC2k 0.9903698 636.9022   2   2 0.001567638
## Average_fixed_raters    ICC3k 0.9984299 636.9022   2   2 0.001567638
## 
##  DNAmTL : 
##                          type       ICC        F df1 df2           p
## Single_raters_absolute   ICC1 0.9898538 196.1186   2   3 0.000661295
## Single_random_raters     ICC2 0.9898538 196.1186   2   2 0.005073088
## Single_fixed_raters      ICC3 0.9898538 196.1186   2   2 0.005073088
## Average_raters_absolute ICC1k 0.9949010 196.1186   2   3 0.000661295
## Average_random_raters   ICC2k 0.9949010 196.1186   2   2 0.005073088
## Average_fixed_raters    ICC3k 0.9949010 196.1186   2   2 0.005073088
## 
##  IEAA : 
##                          type       ICC        F df1 df2           p
## Single_raters_absolute   ICC1 0.9750351 79.11262   2   3 0.002538238
## Single_random_raters     ICC2 0.9750756 90.94104   2   2 0.010876536
## Single_fixed_raters      ICC3 0.9782469 90.94104   2   2 0.010876536
## Average_raters_absolute ICC1k 0.9873598 79.11262   2   3 0.002538238
## Average_random_raters   ICC2k 0.9873805 90.94104   2   2 0.010876536
## Average_fixed_raters    ICC3k 0.9890039 90.94104   2   2 0.010876536
## 
##  EEAA : 
##                          type       ICC        F df1 df2          p
## Single_raters_absolute   ICC1 0.8698274 14.36421   2   3 0.02907429
## Single_random_raters     ICC2 0.8698274 14.36421   2   2 0.06508632
## Single_fixed_raters      ICC3 0.8698274 14.36421   2   2 0.06508632
## Average_raters_absolute ICC1k 0.9303825 14.36421   2   3 0.02907429
## Average_random_raters   ICC2k 0.9303825 14.36421   2   2 0.06508632
## Average_fixed_raters    ICC3k 0.9303825 14.36421   2   2 0.06508632

Remove duplicated observations

Number of observations removed: 3

Missing values

For each exposure measurements:

Number of missing values in baseline covariates:

##             Age             BMI predictedGender        fueltype       stovetype 
##               0               0               0               0               0

Number of missing values in ambient exposure measurements covariates:

##  bap_air pm25_air  ANY_air  BPE_air  BaA_air  BbF_air  BkF_air  CHR_air 
##        5        0       41        5        5        5        5        5 
##  DBA_air  FLT_air  FLU_air  IPY_air  NAP_air  PHE_air  PYR_air 
##        5        5       41        5       41       41        5

Number of missing values in urinary measurements covariates:

## Benzanthracene_Chrysene_urine             Naphthalene_urine 
##                             2                             0 
##     Methylnaphthalene_2_urine     Methylnaphthalene_1_urine 
##                             9                             4 
##            Acenaphthene_urine Phenanthrene_Anthracene_urine 
##                             0                             0 
##            Fluoranthene_urine                  Pyrene_urine 
##                             0                            18

For each observations:

Observations with different number of total missing measurements:

##                                                       
## Number of missing measurements:  0  1 2 3  4 5 6 14 15
## Number of observations:         73 10 4 1 27 8 1  3  2

Observations with different number of missing air exposure measurements:

##                                         
## Number of missing measurements:  0  4 14
## Number of observations:         88 36  5

Observations with different number of missing urinary measurements:

##                                           
## Number of missing measurements:   0  1 2 3
## Number of observations:         103 20 5 1

Number of observations removed: 0

Table with summarize info [table 1]

Characteristic	N = 129¹
Age	58 (15)
BMI	22.0 (3.5)
predictedGender
female	129 / 129 (100%)
fueltype
Plant	4 / 129 (3.1%)
Smokeles	18 / 129 (14%)
Smoky	98 / 129 (76%)
Wood	9 / 129 (7.0%)
stovetype
FP	11 / 129 (8.5%)
HS	31 / 129 (24%)
LS	30 / 129 (23%)
Mix	37 / 129 (29%)
PS	20 / 129 (16%)
DNAmAge	59 (13)
DNAmAgeHannum	60 (14)
DNAmPhenoAge	56 (14)
DNAmAgeSkinBloodClock	57 (13)
DNAmGrimAge	56 (11)
DNAmTL	6.82 (0.32)
AgeAccelerationResidual	-0.1 (4.4)
AgeAccelerationResidualHannum	-0.1 (4.0)
AgeAccelPheno	0.0 (4.3)
DNAmAgeSkinBloodClockAdjAge	0.0 (3.4)
AgeAccelGrim	-0.09 (2.99)
DNAmTLAdjAge	0.00 (0.18)
IEAA	0.0 (4.1)
EEAA	-0.1 (5.1)
bap_air	65 (87)
(Missing)	5
pm25_air	197 (176)
ANY_air	1,038 (1,792)
(Missing)	41
BPE_air	71 (92)
(Missing)	5
BaA_air	88 (145)
(Missing)	5
BbF_air	106 (143)
(Missing)	5
BkF_air	23 (32)
(Missing)	5
CHR_air	83 (133)
(Missing)	5
DBA_air	24 (36)
(Missing)	5
FLT_air	60 (136)
(Missing)	5
FLU_air	486 (680)
(Missing)	41
IPY_air	43 (51)
(Missing)	5
NAP_air	5,731 (8,737)
(Missing)	41
PHE_air	730 (1,054)
(Missing)	41
PYR_air	66 (139)
(Missing)	5
Benzanthracene_Chrysene_urine	0.92 (3.17)
(Missing)	2
Naphthalene_urine	224 (686)
Methylnaphthalene_2_urine	46 (59)
(Missing)	9
Methylnaphthalene_1_urine	20 (24)
(Missing)	4
Acenaphthene_urine	7 (11)
Phenanthrene_Anthracene_urine	204 (275)
Fluoranthene_urine	21 (23)
Pyrene_urine	0.71 (0.57)
(Missing)	18
¹ Mean (SD); n / N (%)

PCA

Explore clocks

Correlogram

Between clocks

Median absolute error

##               DNAmAge         DNAmAgeHannum          DNAmPhenoAge 
##              3.391653              3.380488              3.005900 
## DNAmAgeSkinBloodClock           DNAmGrimAge 
##              2.585191              4.063720

Scatterplots of age vs each epigenetic age

Explore Exposure Measurements

Correlogram

Between Ambient Exposure Measurements

Between Urinary Exposure Measurements

Between Ambient and Urinary Exposure Measurements

Clocks vs Measurements

Pairplots

New data of fuelExp

Childhood Exposure

All subjects

Characteristic	N = 224¹
brthFuel
Mix	101 / 220 (46%)
Smokeles	9 / 220 (4.1%)
Smoky	78 / 220 (35%)
Wood	32 / 220 (15%)
(Missing)	4
¹ n / N (%)

subjects with DNA methylation data

Characteristic	N = 129¹
brthFuel
Mix	54 / 126 (43%)
Smokeles	7 / 126 (5.6%)
Smoky	52 / 126 (41%)
Wood	13 / 126 (10%)
(Missing)	3
¹ n / N (%)

Lifetime Exposure

All subjects

Characteristic	N = 224¹
cumFuel
Mix	168 / 224 (75%)
Smoky	56 / 224 (25%)
¹ n / N (%)

subjects with DNA methylation data

Characteristic	N = 129¹
cumFuel
Mix	87 / 129 (67%)
Smoky	42 / 129 (33%)
¹ n / N (%)

Fueltype from our original dataset

All subjects

Characteristic	N = 224¹
fueltype
Plant	4 / 129 (3.1%)
Smokeles	18 / 129 (14%)
Smoky	98 / 129 (76%)
Wood	9 / 129 (7.0%)
(Missing)	95
¹ n / N (%)

subjects with DNA methylation data

Characteristic	N = 224¹
fueltype
Plant	4 / 129 (3.1%)
Smokeles	18 / 129 (14%)
Smoky	98 / 129 (76%)
Wood	9 / 129 (7.0%)
(Missing)	95
¹ n / N (%)

Control Subjects

## # A tibble: 6 × 14
##   ids_LCW VISIT education_cat SbjctD_LEX loc_N loc_E grp            dist_min
##   <chr>   <lgl>         <dbl> <lgl>      <dbl> <dbl> <chr>             <dbl>
## 1 1824    NA                3 NA          26.2  104. LCW Clean Ctrl    0.347
## 2 1912    NA                5 NA          26.2  104. LCW Clean Ctrl    0.347
## 3 2276    NA                6 NA          26.2  104. LCW Clean Ctrl    0.347
## 4 4386    NA                2 NA          25.7  104. LCW Clean Ctrl    0.570
## 5 2394    NA                2 NA          26.2  104. LCW Clean Ctrl    0.607
## 6 2274    NA                5 NA          26.2  104. LCW Clean Ctrl    0.609
## # … with 6 more variables: inpatient_sms <dbl>, diagnosis_code_sms <dbl>,
## #   enroll_age_sms <dbl>, diagnosis_eng_sms <chr>, econ_any_items <dbl>,
## #   curr_commune <chr>

Exploration

Seraphina Shi

Load Data from the Clock Calculator

Duplicated Observations

Clocks stable?

Concordance between pairs

Absolute differences in two observations

ICC of each clock

Remove duplicated observations

Missing values

For each exposure measurements:

For each observations:

Table with summarize info [table 1]

PCA

Explore clocks

Correlogram

Between clocks

Median absolute error

Scatterplots of age vs each epigenetic age

Explore Exposure Measurements

Correlogram

Between Ambient Exposure Measurements

Between Urinary Exposure Measurements

Between Ambient and Urinary Exposure Measurements

Clocks vs Measurements

Pairplots

New data of fuelExp

Childhood Exposure

All subjects

subjects with DNA methylation data

Lifetime Exposure

All subjects

subjects with DNA methylation data

Fueltype from our original dataset

All subjects

subjects with DNA methylation data

Control Subjects